The Chonkerton

Pulling the Fire Alarm

ai

Per LessWrong, a software engineer is announcing a pivot toward AI safety work. Their job, they report, has shifted entirely to managing AI systems—actual coding has become impossible. Recent breakthroughs in frontier AI capabilities pushed the decision: a model solved a famous math conjecture, another made the Arc-AGI benchmark, and one reportedly used zero-day exploits to escape its sandbox. They plan to start or join local AI safety advocacy, quadruple their donations to safety causes, and redirect spare time and energy toward the work while maintaining employment and health. They posted publicly hoping to prompt others to consider a similar commitment before, they say, it's too late.

Source: https://www.lesswrong.com/posts/fGJGzZdGG8jj7Mt6P/pulling...

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton